AAS 247: Photometric Catalog Integration in a Python Framework for Spectral Energy Distribution Construction and Young Stellar Object Classification

~ newton ~ Leave a comment

Abstract

One method for studying young stellar objects (YSOs) is by assembling their spectral energy distributions (SEDs) from multi-wavelength photometric catalogs. These SEDs can be used to estimate the relative ages of YSOs. In the Cepheus C region, we took an existing photometric catalog created using both infrared and visible data from 2MASS, Spitzer, WISE, Herschel, SDSS, IPHAS, and PanSTARRS missions, and updated the catalog using data from the Gaia survey. We focused on constructing SEDs and using their shapes for preliminary classification. To support this effort, we developed a Python-based Google Colab notebook that implements SED construction using tools such as Astropy, Astroquery, and pandas. While generating SED plots can be computationally straightforward with a completed notebook, verifying catalog matches and cleaning legacy data is labor-intensive and critical to data integrity. The notebook is meant as a part of the process of teaching students to access, manipulate, visualize, and explain astronomical datasets. Our pipeline is designed to help students engage with real astronomical data, emphasizing transparency in data handling and reproducibility. The use of a computational essay format allows for narrative text and code blocks to co-exist in a single structure. This work is part of a broader educational initiative to involve high school students in authentic research through computational essays in Google Colab. We followed best practices in data science education, including the statistical problem-solving process, and structured our notebook to be accessible and reusable for other teams. Preliminary results from Cepheus C will be shared, along with suggestions for extending the catalog with time-series data to explore YSO variability, contributing to the larger YOSVAR project.
iPoster link

Data Driven Amateur Astronomy

~ newton

Houston Astronomical Society December 5th, 2025

Modern astronomy research has become data-driven. Using data science techniques alongside computation allows us to interrogate data to understand astrophysical phenomena. The explosion of data sets has opened up new ways for enterprising amateur astronomers to contribute to modern astronomical research. Data can come from large-scale surveys, space-based observatories, individual scientists, or students. You can learn to select, reduce, visualize, and interpret authentic astronomical data while applying data science techniques to construct astronomy knowledge. Many free web-based tools leverage data science techniques. This talk explores how these activities bridge the gap between data science and astronomy, enabling amateurs to learn about both simultaneously.

The content of this talk can be cited as: Newland, J. (2025). Using Data Science in High School Astronomy. ASP 2024: Astronomy Across the Spectrum, 539, 147. http://arxiv.org/abs/2501.04856

The Google Colab (Jupyter Notebook) developed by Sara Kannan and me can be found here. Note that the actual catalog we created is not publicly available, so this notebook requires an existing catalog for SED creation.

If you are interested in data-driven astronomy learning, check out the page below from a talk given at the first-ever Data Science Education in K12 Conference. Even though the materials shared were designed for teaching high school astronomy, enterprising amateur astronomers can still pick up some cool tricks.

Integrating Data Science Pedagogy into Astronomy Courses

HOU’s Coding: Using CS for Science Teaching

~ newton

Integrating Computer Science into Science Teaching

When doing domain-specific programming in science, some CS pedagogy can be used to
scaffold concepts like conditionals, function writing, and looping. Using worked examples,
minimally working programs, sub-task labeling, and live coding can help a student bring
coding to bear on learning concepts in science. Room B107 1:15 – 2:15 pm CDT

WeTeach_CS Summit 2025

~ newton

Impacting Student Attitudes About CS Through Coding Integration into Science

In this session, participants will learn how integrating computer science (CS) pedagogy into science and math classes can positively impact students’ attitudes about computer science. When students experience lessons in science and math classes that incorporate computer science (CS), they feel more confident about CS. They are more likely to pursue academic pathways that include CS. Participants will see examples of specific lessons that can be used in science and math courses to improve students’ attitudes about the use of computer science (CS) in non-CS STEM courses.

Participants will use STEMcoding during the session and can access STEMcoding for free for one year.

If you are interested in more academic details about the STEMcoding study, click here.

TSAAPT/TSAPS Spring 2025: Modeling with STEMcoding in AP Physics 1

~ newton

Improving Student Computational Thinking Outcome Expectancy through Integrating Model-building in a Secondary Physics Course

The role of computation in physics is increasing in both a professional and educational context. High school physics pedagogy should incorporate computational thinking, data science, and computer science skills to bolster learning across these domains. The study described here shows that when students build physically meaningful models using computer programming in a physics course, their outcome expectancy is positively impacted. Outcome expectancy can predict students’ future academic and professional choices. The STEMcoding platform asks students to construct models using Euler-Cromer step-wise modeling of physical phenomena, such as those described by the laws of motion. The study presented had students use STEMcoding during the kinematics portion of Advanced Placement Physics 1 classes at a large, urban high school in southeast Texas. A moderate positive impact on outcome expectancy was found across various groups in the study population. Results show how computational essays can provide scaffolded learning while allowing students to use code to create interactive models, allowing knowledge construction in a domain-specific manner.